• Korean Journal of Remote Sensing
• /
• v.37 no.1
• /
• pp.111-122
• /
• 2021

Building footprint extraction is an active topic in the domain of remote sensing, since buildings are a fundamental unit of urban areas. Deep convolutional neural networks successfully perform footprint extraction from optical satellite images. However, semantic segmentation produces coarse results in the output, such as blurred and rounded boundaries, which are caused by the use of convolutional layers with large receptive fields and pooling layers. The objective of this study is to generate visually enhanced building objects by directly extracting the vertices of individual buildings by combining instance segmentation and keypoint detection. The target keypoints in building extraction are defined as points of interest based on the local image gradient direction, that is, the vertices of a building polygon. The proposed framework follows a two-stage, top-down approach that is divided into object detection and keypoint estimation. Keypoints between instances are distinguished by merging the rough segmentation masks and the local features of regions of interest. A building polygon is created by grouping the predicted keypoints through a simple geometric method. Our model achieved an F1-score of 0.650 with an mIoU of 62.6 for building footprint extraction using the OpenCitesAI dataset. The results demonstrated that the proposed framework using keypoint estimation exhibited better segmentation performance when compared with Mask R-CNN in terms of both qualitative and quantitative results.

• Journal of Korean Institute of Architectural Sustainable Environment and Building Systems
• /
• v.12 no.6
• /
• pp.579-590
• /
• 2018

Anomaly detection on building energy consumption has been regarded as an effective tool to reduce energy saving on building operation and maintenance. However, it requires energy model and FDD expert for quantitative model approach or large amount of training data for qualitative/history data approach. Both method needs additional time and labors. This study propose a machine learning and data science approach to define faulty conditions on hourly building energy consumption with reducing data amount and input requirement. It suggests an application of Support Vector Data Description (SVDD) method on training normal condition of hourly building energy consumption incorporated with hourly outdoor air temperature and time integer in a week, 168 data points and identifying hourly abnormal condition in the next day. The result shows the developed model has a better performance when the ${\nu}$ (probability of error in the training set) is 0.05 and ${\gamma}$ (radius of hyper plane) 0.2. The model accuracy to identify anomaly operation ranges from 70% (10% increase anomaly) to 95% (20% decrease anomaly) for daily total (24 hours) and from 80% (10% decrease anomaly) to 10%(15% increase anomaly) for occupied hours, respectively.

• Journal of the Society of Disaster Information
• /
• v.18 no.2
• /
• pp.364-373
• /
• 2022

Purpose: The purpose of this paper is to develope smoke detection using AI model for detecting the initial fire in underground utility tunnels using CCTV Method: To improve detection performance of smoke which is high irregular, a deep learning model for fire detection was trained to optimize smoke detection. Also, several approaches such as dataset cleansing and gradient exploding release were applied to enhance model, and compared with results of those. Result: Results show the proposed approaches can improve the model performance, and the final model has good prediction capability according to several indexes such as mAP. However, the final model has low false negative but high false positive capacities. Conclusion: The present model can apply to smoke detection in underground utility tunnel, fixing the defect by linking between the model and the utility tunnel control system.

• Smart Structures and Systems
• /
• v.6 no.7
• /
• pp.811-833
• /
• 2010

Vibration control and health monitoring of building structures have been actively investigated in recent years but treated separately according to the primary objective pursued. This paper presents a general approach in the time domain for integrating vibration control and health monitoring of a building structure to accommodate various types of control devices and on-line damage detection. The concept of the time-domain approach for integrated vibration control and health monitoring is first introduced. A parameter identification scheme is then developed to identify structural stiffness parameters and update the structural analytical model. Based on the updated analytical model, vibration control of the building using semi-active friction dampers against earthquake excitation is carried out. By assuming that the building suffers certain damage after extreme event or long service and by using the previously identified original structural parameters, a damage detection scheme is finally proposed and used for damage detection. The feasibility of the proposed approach is demonstrated through detailed numerical examples and extensive parameter studies.

• International conference on construction engineering and project management
• /
• 2022.06a
• /
• pp.893-903
• /
• 2022

Recently, computer vision (CV)-based safety monitoring (i.e., object detection) system has been widely researched in the construction industry. Sufficient and high-quality data collection is required to detect objects accurately. Such data collection is significant for detecting small objects or images from different camera angles. Although several previous studies proposed novel data augmentation and synthetic data generation approaches, it is still not thoroughly addressed (i.e., limited accuracy) in the dynamic construction work environment. In this study, we proposed a game engine-driven synthetic data generation model to enhance the accuracy of the CV-based object detection model, mainly targeting small objects. In the virtual 3D environment, we generated synthetic data to complement training images by altering the virtual camera angles. The main contribution of this paper is to confirm whether synthetic data generated in the game engine can improve the accuracy of the CV-based object detection model.

• Structural Engineering and Mechanics
• /
• v.26 no.3
• /
• pp.315-341
• /
• 2007

A sudden change of stiffness in a structure, associated with the events such as weld fracture and brace breakage, will cause a discontinuity in acceleration response time histories recorded in the vicinity of damage location at damage time instant. A new damage index is proposed and implemented in this paper to detect the damage time instant, location, and severity of a structure due to a sudden change of structural stiffness. The proposed damage index is suitable for online structural health monitoring applications. It can also be used in conjunction with the empirical mode decomposition (EMD) for damage detection without using the intermittency check. Numerical simulation using a five-story shear building under different types of excitation is executed to assess the effectiveness and reliability of the proposed damage index and damage detection approach for the building at different damage levels. The sensitivity of the damage index to the intensity and frequency range of measurement noise is also examined. The results from this study demonstrate that the damage index and damage detection approach proposed can accurately identify the damage time instant and location in the building due to a sudden loss of stiffness if measurement noise is below a certain level. The relation between the damage severity and the proposed damage index is linear. The wavelet-transform (WT) and the EMD with intermittency check are also applied to the same building for the comparison of detection efficiency between the proposed approach, the WT and the EMD.

• Structural Engineering and Mechanics
• /
• v.30 no.4
• /
• pp.445-466
• /
• 2008

This paper presents a novel structural damage detection method with a new damage index based on the statistical moments of dynamic responses of a structure under a random excitation. After a brief introduction to statistical moment theory, the principle of the new method is put forward in terms of a single-degree-of-freedom (SDOF) system. The sensitivity of statistical moment to structural damage is discussed for various types of structural responses and different orders of statistical moment. The formulae for statistical moment-based damage detection are derived. The effect of measurement noise on damage detection is ascertained. The new damage index and the proposed statistical moment-based damage detection method are then extended to multi-degree-of-freedom (MDOF) systems with resort to the leastsquares method. As numerical studies, the proposed method is applied to both single and multi-story shear buildings. Numerical results show that the fourth-order statistical moment of story drifts is a more sensitive indicator to structural stiffness reduction than the natural frequencies, the second order moment of story drift, and the fourth-order moments of velocity and acceleration responses of the shear building. The fourth-order statistical moment of story drifts can be used to accurately identify both location and severity of structural stiffness reduction of the shear building. Furthermore, a significant advantage of the proposed damage detection method lies in that it is insensitive to measurement noise.

• Journal of the Korean Society of Surveying, Geodesy, Photogrammetry and Cartography
• /
• v.17 no.4
• /
• pp.421-429
• /
• 1999

It is estimated that, in the twenty-first century, 70％ of global citizens will live in urban areas. This accelerated urbanization will require a greater need for the building DEM and orthoimagery along with Geographic Information System for urban management. The building DEM requires the detection of outlines showing building shapes. To do this, automatic and semiautomatic building extractions are usually used. However, in cases where automatic extraction is performed directly from the aerial images, accurate building outline extraction is very difficult because of shadow, roof color, and neighboring trees making it hard to discern building roofs. To overcome this problem semiautomatic building extraction was suggested in this paper. When a roof texture was homogeneous, building outline detection was performed by mouse-clicking on a part of the roof. To construct the building outlines when the texture was not homogeneous, a computer program was developed to search out corner points by clicking spots near corner points. The building DEM was generated by taking into account building outlines and heights calculated by image matching.

• Journal of the Korean Society of Surveying, Geodesy, Photogrammetry and Cartography
• /
• v.40 no.4
• /
• pp.359-364
• /
• 2022

The change detection in remote sensing results in errors due to the camera's optical factors, seasonal factors, and land cover characteristics. The inclination of the building in the image was simulated according to the camera angle using the Cycle Generative Adversarial Network method, and the simulated image was used to contribute to the improvement of change detection accuracy. Based on CycleGAN, the inclination of the building was similarly simulated to the building in the other image based on the image of one of the two periods, and the error of the original image and the inclination of the building was compared and analyzed. The experimental data were taken at different times at different angles, and Kompsat-3A high-resolution satellite images including urban areas with dense buildings were used. As a result of the experiment, the number of incorrect detection pixels per building in the two images for the building area in the image was shown to be reduced by approximately 7 times from 12,632 in the original image and 1,730 in the CycleGAN-based simulation image. Therefore, it was confirmed that the proposed method can reduce detection errors due to the inclination of the building.

• International conference on construction engineering and project management
• /
• 2022.06a
• /
• pp.136-144
• /
• 2022

Construction sites are characterized by dangerous situations and environments that cause fatal accidents. Potential risk detection needs to be improved by continuously monitoring site conditions. However, the current labor-intensive inspection practice has many limitations in monitoring dangerous conditions at construction sites. Computer vision technology that can quickly analyze and collect site conditions from images has been in the spotlight as a solution. Nonetheless, inspection results obtained via computer vision are still stored and managed in centralized systems vulnerable to tampering with information by the central node. Blockchain has been used as a reliable and efficient decentralized information management system. Despite its potential, only limited research has been conducted integrating computer vision and blockchain. Therefore, to solve the current safety management problems, the authors propose a framework for construction site inspection that integrates object detection and blockchain network, enabling efficient and reliable remote inspection. Object detection is applied to enable the automatic analysis of site safety conditions. As a result, the workload of safety managers can be reduced with inspection results stored and distributed reliably through the blockchain network. In addition, errors or forgery in the inspection process can be automatically prevented and verified through a smart contract. As site safety conditions are reliably shared with project participants, project participants can remotely inspect site conditions and make safety-related decisions in trust.